TL;DR: The floral morphology of Miconia brevitheca and M. tonduzii suggest these species are generalist pollinated, adding support to previous suggestions of a shift from pollinator specialization to generalization.
Abstract: We report observations of generalist insect visitation and probable pollination of the flowers of Miconia brevitheca and M tonduzii var tonduzii, both of which belong to Miconia section Cremanium The most diverse assemblage of insect visitors was recorded for M brevitheca, which has some of the smallest flowers in the Melastomataceae Visitors included butterflies of at least five families Miconia tonduzii, on the other hand, was visited mostly by Diptera of the families Syrphidae and Tachinidae Scanning electron microscopy revealed stomata on the lower part of the anthers on the abaxial surface, which has been recently reported in other melastomes as the nectar-secreting site Flowers of a closely related buzz pollinated control, Leandra subseriata, confirmed the absence of stomata in the flowers of this species Flowers of these Miconia also are unusual in that they have broad pores, filaments that are geniculate at the middle of their length, and short styles that are not exserted beyond the stamens (not herkogamous) The floral morphology, along with observations of many generalist insect visitors and almost no buzzing bees, suggest these species are generalist pollinated The implications of these observations are relevant to the understanding of pollinator systems, because the lineage to which these species belong, Miconia section Cremanium, is nested in a bee-pollinated clade, adding support to previous suggestions of a shift from pollinator specialization to generalization